RMS Surface Roughness: What Rq Means and How to Measure It

RMS surface roughness is a numerical description of height variation in a surface profile. It is usually identified as Rq: the root mean square deviation of the assessed roughness profile. The calculation gives extra weight to larger deviations because each one is squared before averaging.

That definition is simple. Using the number correctly is not. An Rq value is meaningful only with its units, measurement direction, evaluation length, filtering rules, instrument method and the standard or procedure used. Change those conditions and the same physical surface can produce a different result.

Quick answer: RMS roughness and Rq usually refer to the same profile parameter. Rq is calculated by squaring the profile-height deviations from the mean line, averaging those squares, and taking the square root. It is not interchangeable with Ra, and there is no universal Ra-to-Rq conversion factor. For purchasing or acceptance, specify and report the complete measurement conditions—not just a number followed by “RMS.”

What Does RMS Surface Roughness Measure?

A measured surface profile contains vertical deviations above and below a reference mean line. After the measurement data have been processed to separate roughness from longer-scale form and waviness, Rq describes the quadratic mean of those roughness-profile deviations over the applicable evaluation length.

In continuous form, the idea is:

Rq = square root of the mean of z(x)² over the evaluation length

Here, z(x) is the profile height relative to the mean line at position x. For discrete data, the instrument squares each sampled height, adds the results, divides by the number of points and takes the square root.

The National Institute of Standards and Technology lists Rq as the root mean square deviation of the assessed profile and distinguishes it from Pq for a primary profile and Wq for a waviness profile (NIST height parameters). The prefix matters: a result calculated from a different profile is not automatically an Rq result.

Rq has the same dimensional unit as the measured heights, commonly micrometres (µm) or microinches (µin). Converting units is exact—1 µm equals approximately 39.37 µin—but converting one roughness parameter into another is not.

Rq vs Ra: The Important Difference

Ra and Rq can be calculated from the same filtered profile, but they summarize it differently.

Rq cannot be lower than Ra when both are calculated from the same set of deviations under identical processing. Equality occurs only when all absolute deviations have the same magnitude, including the trivial perfectly flat case. Real profiles normally contain a distribution of heights, so Rq is usually higher.

The size of that difference depends on the distribution of the profile heights. A ratio derived for an ideal sine wave or an assumed statistical distribution is not a general conversion factor for turned, ground, honed, blasted, coated or worn surfaces. Estimating “RMS = Ra × 1.11” may produce a plausible-looking number that does not represent the measured surface.

A Small Calculation Example

Consider five simplified height deviations, in micrometres, after establishing a mean line:

−2, −1, 0, +1, +2

For Ra, take the absolute values and average them:

Ra = (2 + 1 + 0 + 1 + 2) / 5 = 1.20 µm

For Rq, square the values, average the squares and take the square root:

Rq = √[(4 + 1 + 0 + 1 + 4) / 5] = √2 = 1.41 µm

Both results describe the same five points, but Rq is higher because the ±2 µm deviations contribute four times as much as the ±1 µm deviations before the final square root. This example illustrates the mathematics only. It is not a standardized measurement, a drawing tolerance or a substitute for a complete profile evaluation.

Why an Rq Number Alone Is Incomplete

A profilometer does not encounter a mathematical roughness profile waiting on the part. It records a mixture of surface form, waviness, roughness, instrument response and noise. Processing choices determine which spatial scales remain in the assessed roughness profile.

This is why two laboratories can report different Rq values without either calculation being arithmetically wrong. Common causes include:

• different nesting indices or cutoff settings;
• different evaluation lengths or numbers of traverses;
• measurements taken parallel rather than perpendicular to the dominant lay;
• different stylus-tip geometry, measuring force or optical objective;
• different form removal, levelling or data treatment;
• different locations on a non-uniform surface;
• contamination, vibration or poor fixturing;
• use of a line-profile parameter for one report and an areal parameter for another.

ISO 21920-2:2021 specifies terms, definitions and parameters for profile-based surface texture. ISO lists the earlier ISO 4287:1997 as withdrawn, so a drawing, procedure or software screen that cites ISO 4287 should not be silently reinterpreted as ISO 21920 (ISO 21920-2:2021). The applicable contract and drawing revision must decide whether a legacy requirement remains controlling or needs formal clarification.

Filtering, Length and Direction Change the Result

Filtering separates surface features by lateral scale. A long-wavelength component may be classified as waviness or form under one setting and remain in the roughness profile under another. ISO 16610-21:2025 specifies linear Gaussian filters for separating large- and small-scale lateral components of surface profiles (ISO 16610-21:2025).

The practical lesson is not to choose a cutoff simply because it makes a trace fit on a short feature. The selected specification operator or approved procedure should control. If the available surface is too short for the required trace, record the limitation and obtain engineering agreement rather than modifying settings without disclosure.

Measurement direction matters too. A turned or ground surface often has a dominant lay. Traversing across the lay may cross many tool marks, while traversing along it may follow a groove and return a very different profile. Record the direction and use the drawing or governing standard to define it.

Repeat measurements at representative locations when texture is not uniform. A compliant average may still hide a local defect if the specification controls individual results.

Profile Rq Is Not Areal Sq

The letter before the parameter identifies the data basis. Rq is derived from an assessed roughness profile along a line. Sq is a root mean square height parameter calculated over a scale-limited area. ISO 25178-2:2021 covers parameters for areal surface-texture methods (ISO 25178-2:2021).

Rq and Sq share a root-mean-square concept, but they are not interchangeable. A line can miss directional features, pores or isolated damage that an area captures; an area measurement introduces its own choices for field size, resolution, levelling, filtering and invalid points. Report the parameter exactly as calculated and do not relabel Sq as Rq because both are described as “RMS roughness.”

Stylus and Optical Measurements Are Not Automatically Equivalent

A contact stylus traces the surface mechanically. Its tip cannot enter every narrow valley, and its force, radius and response affect what it records. Optical results depend on the method, reflectivity, slope, focus, resolution and data-reconstruction rules.

NIST has documented both mechanical and optical surface-topography methods and emphasizes calibration, reference materials and measurement uncertainty (NIST surface texture programme). In a comparison study, NIST researchers found that optical and stylus methods could agree reasonably on some moderately rough surfaces while discrepancies appeared in lower roughness ranges; method equivalence therefore needs validation for the surface and range in question (NIST method comparison).

Do not approve a supplier-method substitution merely because both instruments display Rq. For acceptance work, establish correlation using representative parts, controlled settings, traceable artifacts and an agreed uncertainty approach.

How to Read an RMS Callout or Inspection Report

Treat “32 RMS” or “0.8 µm RMS” as the start of a question, not a complete requirement. Use this decoding sequence:

1. Identify the parameter. Confirm that “RMS” means profile Rq, not areal Sq, electrical RMS, or an instrument-specific field.
2. Confirm the units. Determine whether the value is in µm, µin or another unit. Never infer units from the size of the number alone.
3. Identify the standard and edition. ISO 21920, ASME B46.1 and legacy drawings can use different rules or defaults. ASME lists B46.1-2019 (R2026) as remaining in effect (ASME B46.1).
4. Read the operator details. Check filtering, nesting indices or cutoffs, evaluation length, direction and any special conditions.
5. Check the decision rule. Determine whether the limit applies to each trace, a mean, a maximum, a population or another defined result.
6. Resolve ambiguity before production. Record an approved interpretation in the purchase order, inspection plan or drawing revision.

For material-sensitive components, surface requirements should be reviewed alongside grade, condition and manufacturing route. AFTO’s stainless steel grades guide explains why an alloy name alone does not complete an engineering specification.

Surface-Roughness Reporting Checklist

Before accepting or comparing a result, confirm that the report states:

• part, feature, batch and measurement-location identification;
• parameter symbol—Rq rather than an unexplained “RMS” label;
• value and unit, with rounding consistent with the requirement;
• governing standard, edition and any customer procedure;
• instrument type, instrument ID and calibration status;
• probe or optical configuration relevant to the method;
• filtering and scale-limitation settings;
• evaluation length or evaluation area;
• trace direction relative to lay;
• number and locations of measurements;
• environmental, cleaning and fixturing conditions where material;
• result distribution, not only a convenient average;
• measurement uncertainty and the acceptance decision rule where required;
• any deviation from the specified method.

If a supplier cannot reproduce the required method or document the settings, treat that as a specification gap. AFTO’s industrial equipment supplier evaluation guide provides a broader framework for making test evidence, deviations and acceptance criteria part of the purchase.

Common RMS Roughness Mistakes

Converting Ra to Rq with a fixed multiplier

Special-case ratios describe assumed profile shapes, not every real surface. Measure Rq or obtain the original profile data and apply the agreed method.

Comparing values with different filters

The formula may be identical while the assessed profiles are not. Match the filtering and evaluation conditions before comparing numbers.

Assuming one average controls defects

Rq responds more strongly to large deviations than Ra, but it still reduces a complex profile to one number. Depending on function, additional height, spacing, material-ratio or defect controls may be necessary.

Treating a visual comparator as an Rq measurement

A comparison specimen can support a workshop judgement when a procedure permits it, but visual or tactile comparison is not the same as a traceable Rq measurement.

Omitting uncertainty near a limit

A displayed value does not prove an unambiguous pass. Close decisions require the specified decision rule and an uncertainty assessment appropriate to the measurement system.

Frequently Asked Questions

Is RMS surface roughness the same as Rq?

Usually, yes, when discussing a line-profile roughness parameter. Confirm the governing standard and software terminology, especially on legacy drawings or areal-measurement reports.

Is Rq always higher than Ra?

For the same finite set of profile deviations processed identically, Rq is greater than or equal to Ra. Real non-flat profiles normally produce a higher Rq. That mathematical relationship does not supply a fixed conversion ratio.

Can I convert microinches RMS to micrometres Rq?

You can convert the unit exactly if the parameter really is the same: multiply µin by 0.0254 to obtain µm. Do not use a unit conversion to disguise uncertainty about whether “RMS” means Rq or whether measurement conditions match.

Which is better, Ra or Rq?

Neither is universally better. Ra is widely used and easy to compare; Rq gives proportionally more influence to larger deviations. The functional requirement, standard and validated measurement method should determine the parameter set.

Final Takeaway

RMS surface roughness is normally the Rq parameter: the root mean square deviation of an assessed roughness profile. Its squared weighting distinguishes it from Ra, but the formula is only one part of the result.

For a defensible decision, identify the parameter and units, apply the correct standard and filtering rules, control direction and location, validate the method, and report relevant uncertainty. A bare “RMS” value is shorthand; a reproducible measurement is a complete system.

Sources

1. International Organization for Standardization. “ISO 21920-2:2021—Surface texture: Profile—Terms, definitions and surface texture parameters.” https://www.iso.org/standard/72226.html
2. International Organization for Standardization. “ISO 25178-2:2021—Surface texture: Areal—Terms, definitions and surface texture parameters.” https://www.iso.org/standard/74591.html
3. International Organization for Standardization. “ISO 16610-21:2025—Linear profile filters: Gaussian filters.” https://www.iso.org/standard/84786.html
4. ASME. “B46.1-2019 (R2026)—Surface Texture (Surface Roughness, Waviness, and Lay).” https://www.asme.org/codes-standards/find-codes-standards/b46-1-surface-texture/2019/pdf
5. National Institute of Standards and Technology. “Height Parameters.” https://physics.nist.gov/VSC/Help/SurfaceParameters/Height_Parameters.jsp
6. National Institute of Standards and Technology. “Surface Texture and Forensic Topography.” https://www.nist.gov/programs-projects/surface-texture-and-forensic-topography
7. Vorburger, T. V., et al. “Comparison of Optical and Stylus Methods for Measurement of Rough Surfaces.” NIST, 2007. https://www.nist.gov/publications/comparison-optical-and-stylus-methods-measurement-rough-surfaces